TS Fantasy NBA: 18/19 settings

Discussion in 'Basketball' started by anthak, Sep 24, 2018.

By anthak on Sep 24, 2018 at 12:30 AM
  1. anthak

    anthak Moderator Staff Member

    Joined:
    Jan 3, 2016
    Messages:
    7,460
    Likes Received:
    2,372
    So far we've had 30 of last years players re-commit, and another one express interest who didnt play last season. We are still yet to hear from 18 of last season's players.

    Following the promotions and relegations from last season, this is how our leagues should look this season, assuming everyone comes back:

    Premier League:
    @anthak @karlos @yumcha @HEAVER @stowie @costak @walesy @bgt2110 @G-Train @ddsaints @Auron @LiQuiD_SiXx

    Division 1:
    @Steve @rodgo @Rev @Lethal @Don_Cottagers @dabombers @Rad_E_Cool @tyze1 @Bearfly @tasho @Lano24 @smokinjoemisiti

    Division 2, Conference 1:
    @Mick @Belmont383 @ViQBoZ @Chris White @HargyBear @nicksnow @The Royals @wateman @Jasmine @port_leschenault @OutKast_au @maggots

    Division 2, Conference 2:
    @headmandude @doolz @Peaches @Sparks @Danners @Len @crossy @Trav00 @macca118 @Jmulldihno @Emes @MACCA2509

    Make sure to confirm you want to keep playing in one of these threads.

    We've got one new player expressed interest too:
    @That KI Guy who has played previously, but not last season.

    I propose we keep the promotions and relegation system the same this season. However if we end up with another league we'll need to adjust it, so we'll wait and see on that.

    DRAFT
    Ive set the draft for Sunday 14th October.
    We usually hold it at 9:30pm on a Sunday night, however its not allowing that timeslot this season. We can have it any time of the day except between 4:45pm and 11pm. I'll set up a poll at some stage, but what are all of your initial thoughts on the best time?

    SCORING
    I would like to propose some minor changes to our scoring. As a SuperCoach site, I find it strange that we don't have a scoring system for NBA that more accurately reflects the impact players have on games of basketball. It has grated me for a fair while that scoring buckets (regardless of efficiency) and volume shooting has such a big influence in our system.

    For our first season, back in 2010, we attempted to come up with a scoring system that reflected the impact players had on games, however there were lots of us that season who weren't happy that wings weren't as relevant as PGs and bigs. So, we adjusted the scoring setup for season 2 to reward stats that wings were getting, and the main outcome of that has been that our scoring rewards volume shooters... Often times volume shooters negatively impact their team in real life, but we value them. This doesn't sit well with me.

    I would like to have a methodology for the values we place on stats, so that there is reasoning behind it all.

    @Rad_E_Cool wrote some awesome comments in a thread back in 2013 about our scoring, and with Walesy's help, we've been able to find them. His comments are a fair bit to take in all at once, so Im not going to post it all here, however I'm going to now propose a new scoring system, some of which is similar to what Rad proposed back in 2013, and I will steal some of his rationale here now to explain some of the reasoning behind the values i have settled on...

    Firstly, this is what our system was last season:

    Field Goals Made (FGM): 1
    Field Goals Missed (FGMI): -1
    Free Throws Made (FTM): 1
    Free Throws Missed (FTMI): -2
    Three Pointers Made (3PM): 2
    Offensive Rebounds (OREB): 2
    Defensive Rebounds (DREB): 1
    Assists (AST): 2
    Steals (STL): 5
    Blocks (BLK): 4
    Turnovers (TO): -2
    Ejections (EJ): -10
    Flagrant Fouls (FF): -4
    Technical Fouls (TF): -3
    Disqualifications (DQ): -5
    Double Doubles (DD): 4
    Triple Doubles (TD): 10
    Quadruple Doubles (QD): 50
    Points (PTS): 1
    Team Wins (TW): 1

    ...
    My proposal is based loosely around possessions. The amount of possessions in each game is pretty similar, given the 48 minutes per game and the 24 second shot clock. The way a team wins a game of basketball is they use their possessions efficiently (shoot well), force their opponent to use them inefficiently (blocks and defense, to force bad shots), and gain more possessions than their opponent (steals, rebounds and taking care of the ball).

    When you have the ball, your possession ends when you get a turnover or shoot the ball (whether you score or miss). I'll cover the shooting aspect later. Following a missed shot, it is most likely the defensive team wins the rebound. Many defensive rebounds are really soft/easy and I've allocated them just one point, however the offensive team has an opportunity to win the ball back by getting an O-board. Therefore an offensive rebound is the reverse of a turnover, one wins a possession, one finishes a possession. It is the same considering steals and turnovers. Therefore, I have allocated turnovers, steals and offensive rebounds all the same relative value, balancing each other out. Furthermore, a block with a defensive rebound has the same outcome as a steal and therefore (given a def reb =1) I have kept a block 1 point less than a steal.

    I have reduced the value for double doubles and triple doubles because they dont really mean anything. The stat that wins the double/triple double is no more valuable than any of the other stats and if a player is good enough to get a double/triple double then they are getting good points for the game anyway, even without the bonus. I have halved their values. The main reasons I propose to keep them at all are because they are widely regarded milestones denoting a good game, and it is fun to watch your player hoping they get to the milestone. I have kept quad dubs at 50 though! because it would be friggin awesome if anyone got there again...

    Fouls are mostly the same. I have reduced the penalty for ejections because I feel -10 was too harsh for a player already getting negative points for the techs/flagrants and missing out on stats from not playing the remainder of the game. In saying that, I do still think there should be a points penalty for getting kicked out of the stadium. Ive increased the penalty for flagrant fouls, mainly because they're far worse than techs. Both are basically for unsportsmanlike conduct not in the spirit of the game, however techs only result in one FT to the opposition, whereas flagrants result in two free throws as well as possession to the opposition, and thus should receive a greater points penalty.

    Before I move onto shooting efficiency, I havent mentioned assists or team wins. I propose to keep assists at 2, they help your team win and are worth less than scoring the basket, so I think its fair. Team wins, I propose to move up to a value of 2. There isnt any other way to measure the players on the winning team who have contributed to the win without any counting stats. Even if they have played poorly, their team has still won, which is the aim of the game, and who knows, maybe they didnt look to play well, but their talk on D helped create a crucial stop that helped win the game... We will never know for sure, but a value of 2 isnt huge anyway.

    Assigning values for shooting was very difficult and took me so long to work out. I tried many different variations of values for attempts, misses, makes and points and have settled somewhere that isnt perfect but is as close as I can get it in my view.

    Last season, league averages for shooting %s were: FG%: 46%, 3P%: 36.2%, FT%: 76.7%, which is all fairly similar to previous seasons too.

    In my proposal, I have set it up so that a player will score 1 fantasy point per actual point at the following %s: 50%2FG, 33.3%3FG, & 75%FT. If they shoot worse than those %s they will score less than 1 fantasy point per point; and if they shoot better then these %s they will score more than 1 fantasy point per point. They will score 0 fantasy points per point at the following %s: 25%2FG, 16.6%3FG, & 60%FT. If they shoot worse than those percentages, they will score negative fantasy points. All of that, I'm fairly happy with, however I couldnt balance it out for if they shoot up near 100%. 100% 2 pointers will score 1.5 fpts per point, 100% 3 pointers will score 1.66 fpts per point, & 100% FTs will score 2 fpts per point. All in all, its actually not that different to how we had it, however I think this will be better. It still benefits 3 point shooters, however not as much.

    Ive ended up going with a system that definitely rewards the more efficient scorers and doesn't suit volume chuckers as much as we had it in previous seasons, but its also still fairly similar. With the increase to offensive rebounds and more harsh penalty for turnovers, it all balances out nicely.

    Here is my proposed scoring system:

    Field Goals Made (FGM): 1
    Field Goals Missed (FGMI): -1
    Free Throws Made (FTM): 1
    Free Throws Missed (FTMI): -3
    Three Pointers Made (3PM): 1
    Offensive Rebounds (OREB): 4
    Defensive Rebounds (DREB): 1
    Assists (AST): 2
    Steals (STL): 4
    Blocks (BLK): 3
    Turnovers (TO): -4
    Ejections (EJ): -5
    Flagrant Fouls (FF): -5
    Technical Fouls (TF): -3
    Disqualifications (DQ): -5
    Double Doubles (DD): 2
    Triple Doubles (TD): 5
    Quadruple Doubles (QD): 50
    Points (PTS): 1
    Team Wins (TW): 2

    .......
    Here is a list of the top 15 players, based on last season's stats and this new scoring system, ranked by average fpts:
    Davis, LeBron, Harden, Giannis, KAT, Russ, AD, Jokic, Curry, LMA, DMC, KD, Lillard, Butler, CP.


    I'm hoping we can just agree on this quickly and move onto setting up the leagues etc, however please all let us know your thoughts and please speak up if you object in any way.

    I have set up a poll, which will only be open for the next three days, until midnight on Wednesday night.
     
    Last edited: Sep 24, 2018
?

Do you agree with the proposed scoring changes?

Poll closed Sep 27, 2018.
  1. Yes, I agree

    7 vote(s)
    33.3%
  2. I prefer to keep it the same as last season

    7 vote(s)
    33.3%
  3. I would prefer a different scoring system (please comment)

    4 vote(s)
    19.0%
  4. I dont mind either way

    3 vote(s)
    14.3%

Comments

Discussion in 'Basketball' started by anthak, Sep 24, 2018.

    1. anthak
      anthak
      hey folks,
      if you vote for the 3rd option of preferring a different scoring system, please comment here in the thread to explain why and what your preferred scoring system is
    2. Len
      Len
      Pretty easy to see you've put a great deal of effort into the proposed revision mate.
      I'm still pretty bubbish here so rely on external info sources quite a bit, the scoring model by nature is very complex and doesn't feel like something we can model easily pre-season.
      I know Rotowires draft forcaster doesn't support half our scoring mechs as it is. Is it possible for you to give us the top dozen or so scorers for fantasy points via scoring alone last year vs the new model?

      Love the direction you want to take it btw
      • Like Like x 2
    3. stowie
      stowie
      Great detail, research and effort @anthak, really appreciate all the effort you put into this every season!

      One for comparisons sake, the official scoring from NBA.com.
      • Points: 1
      • Rebounds: 1.2
      • Assists: 1.5
      • Steals: 3
      • Blocks: 3
      • Turnovers: -1
      It's pretty simple, using only the main categories, but having a quick look over some players the end result in FP are quite similar.

      Player - New Scoring / Official Scoring
      LBJ - 55.5 / 54.1
      Giannis - 54.7 / 51.7
      Klay - 32.3 / 30.3
      Drummond - 52.5 / 45.5
      AD - 63.2 / 55.1
      CJ - 34 / 33.5

      A full list using the official scoring can be found here.

      For me, one of the big advantages is being able to take in more fantasy articles, podcasts, etc as they all would become relevant to our system.
      • Like Like x 2
    4. anthak
      anthak
      Thanks for the suggestion Stowie.

      I agree that there would be an advantage to being more mainstream given the analysis provided by many on the podcasts articles etc, however I reckon we can do better than that system.

      My main problem with the "official" scoring system is that it doesnt regard shooting efficiency at all. You get a point for every point scored, regardless of how many shots put up.
      However I also dont like how offensive rebounds are treated the same as defensive rebounds as mentioned in the OP, and also that -1 for turnovers isnt harsh enough.
      I'm also not a fan of blocks being equal to steals, for reasons outlined in the OP, but also because some analysts have found that blocks have very little effect on the outcome of games (John Hollinger is one who comes to mind, even though I dont necessarily like the PER or Game Score metrics either, for other reasons)
    5. anthak
      anthak
      These are public leagues so I think this should work...
      Check these two links:
      Premier league is currently set up using the proposed system: http://fantasy.espn.com/basketball/players/add?leagueId=23530&statSplit=lastSeason
      Div 2 is currently still set at last years system: http://fantasy.espn.com/basketball/players/add?leagueId=10201&statSplit=lastSeason

      Top 15 last season (proposed scoring system rank & ave in brackets):
      1. Harden 70.4 (3rd, 55)
      2. Davis 69.5 (1st, 63.2)
      3. LeBron 69.1 (2nd, 55.5)
      4. Westbrook 66 (6th, 52.7)
      5. DMC 63.8 (11th, 49.5)
      6. Giannis 62.6 (4th, 54.7)
      7. Curry 60.3 (9th, 50.1)
      8. Durant 59.1 (12th, 48.8)
      9. KAT 57.4 (5th, 54)
      10. Lillard 57 (13th, 48.2)
      11. Jokic 56.1 (8th, 50.6)
      12. Drummond 53.6 (7th, 52.5)
      13. Paul 53.3 (15th, 45.8)
      14. Oladipo 52.7 (18th, 42.4)
      15. Butler 52.6 (14th, 47.6)

      Also, LaMarcus Aldridge was ranked 20th last season with an ave of 49.3, but would replace Oladipo in the top 15 under the new system, with his average going up to 49.8.
      • Like Like x 1
    6. Len
      Len
      Unfortunately I'm getting a permissions error.
      Fair to say that's a significant ranking impact at the pointy end
    7. Rad_E_Cool
      Rad_E_Cool
      I definitely agree with changing the system and do like rewarding those that help the team the most, so even though I voted that I would prefer a different scoring system, I would be okay with this change.

      Whilst I appreciate @stowie bringing it back to the NBA.com official system, the rationale for this system was to try and make it engaging for new fans and minimising negative stats - there's probably a whole bunch of gambling studies about how important it was to focus on positive scoring and minimising complexity. I don't think that this works for the a site called Too Serious though.

      My preference is for some of the stats we currently keep be done away with and not afforded any values. In particular, DQs, DDs, TDs, QDs, and EJs. The double-figure stats ones are fun, but not informative, and ejections and DQs hurt the player through no more stats being able to be received, so no need for the extra penalty, particularly as for EJs they would also pick up 2 techs, and 1 or 2 flagrants.

      So a simple counter-proposal would be the following (changes in bold):
      Field Goals Made (FGM): 1
      Field Goals Missed (FGMI): -1
      Free Throws Made (FTM): 1
      Free Throws Missed (FTMI): -3
      Three Pointers Made (3PM): 1
      Offensive Rebounds (OREB): 4
      Defensive Rebounds (DREB): 1
      Assists (AST): 2
      Steals (STL): 4
      Blocks (BLK): 3
      Turnovers (TO): -3
      Ejections (EJ): 0
      Flagrant Fouls (FF): -5
      Technical Fouls (TF): -3
      Disqualifications (DQ): 0
      Double Doubles (DD): 0
      Triple Doubles (TD): 0
      Quadruple Doubles (QD): 0
      Points (PTS): 1
      Team Wins (TW): 2

      -------
      However, I could still suggest further changes based on adjusting the shooting. Whether that be getting rid of makes and misses for the different attempts, to just using attempts with FGA = 2FTA in terms of one possession.

      With your proposal @anthak , I think you've tried your best to get it to reflect efficient shooting, but just thinking about a scenario where someone dunks on another player and draws the foul but misses the and-one opportunity, they would score 2 points plus 1 FGM plus FTMiss = 2+1-3 = 0. Ooof! I mean, I know we can't isolate and-one FTs from other FTs, but I think the FTMiss value is too high. A Hack-a-Howard unsuccessful trip to the line is -6, or 1/2 is -1. This more than the value should be for using up a team possession and coming back with 0 or 1 point (given that 1 point is an average PPP).

      The point is, I have always thought that FTs are an efficient way of scoring and there's no need to try to normalise it to be the same level of efficiency as shooting a field goal. 50% FT shooting is 1 PPP in reality, just like 50% FG and 33% 3FG - I don't see why we have to try and tie it to league averages so much.

      With this, a simple change is to alter the above so that FTMiss = -2.

      A more complex change would be to overhaul it so we only factor in FGAs, FTAs and points scored for each possession - if there's some desire for this I will do something up.
      • Like Like x 3
    8. anthak
      anthak
      must be due to the new website. In previous seasons all our leagues were viewable by anyone.
    9. anthak
      anthak
      Thanks for feedback/suggestions.

      Its a fair point you make re FTs but they should be scored anyway. If players miss them then they deserve to be penalised...

      I dont have my spreadsheets with me right now so I cant check out the ramifications for changing to -2 quickly right now.

      Also, I tried out heaps of various scenarios of using FGAs instead of missed and made shots, but it didnt work as nicely. Id have to check my spreadsheets, but from memory it meant the high % shooting scored too much, like 2 fpts per actual point for 100%FG, rather than 1.5/1.66 fpts per actual point for instance.
    10. Rad_E_Cool
      Rad_E_Cool
      Yeah I figured you'd done the work - don't worry about the FGA vs FGMiss issue then.
      Good work with putting it all together. I think we're always going to disagree on some of these aspects, but your proposed changes are on the same lines as how I'd like to see it, so it's all good.
      • Like Like x 1
    11. dabombers
      dabombers
      Hey Everyone,
      Been a bit quiet, a massive year here so too busy to post and comment on things but i'm still around.
      keen again for both keeper and normal league... Always striving to get back into premier league!!! Thought I drafted the team last year to do such but thems the brakes..
      Putting my two bob in, I think we either keep the scoring the same as it has been for several years or we over hall it to one of the standardized systems that most Fantasy crews use...
      I think we have a great scoring system where no positional players are hurt overall. PG's that rebound Bonus!! Centre's that shoot 3's great, "now that they all wanna do it"!!!
      But I am not for penalising free-throws missed beyond most standard league scoring systems...
      Also Blocks vs Steals vs Assists in the newish system tailers teams to chose everyone bar Centres.
      Centres need to keep there relevancy in the game, and most yet not all have shithouse free throw stats. leaving only a limited few to be fantasy relevant.
      I still reckon a Block and an Offensive rebound in a Game is far more worthwhile than a turnover as the guys who turnover more have the ball in hand more...….
      The guys who hustle and get teams to win are the guys who assist/block/steal and rebound and these stats should be rewarded.. The same as a ball hog who scores 30+ points.. eg Harden...…
      I would like a scoring system that sorts out this but I think we all need to chip in and do a season of analysis to confirm any changes..
      Doing Rash changes a few weeks before opening will cause Draft Chaos!!!
    12. dabombers
      dabombers
      Everyone's Draft Strategies over the years has been to try and get 2 top centres 2 top point guards and if we are lucky with draft positions we might get the lucky wings SG/SF and PF that create a winning team...
      And that is all any NBA coach could ask for... Isn't that what we are trying to replicate???
    13. anthak
      anthak
      Hey man, good to hear from you. I think you may have misunderstood the changes, because this scoring system I’ve proposed favours bigs more than others because they shoot high FG% and get O-boards.

      I’m gonna check out the FT% ramifications later when I get a proper chance.
    14. Steve
      Steve
      Wow Ant you have put a lot of work into this.

      I agree that we do need a bit of tinkering to the scoring system.

      We can definitely do away with the player losing points when they foul out when they are deliberately fouling to keep their team in the game in the last minute of the match. They are doing the right thing by their team in this instance.

      I might be on my lonesome here but I think we should award the same amount of points for TD's as they are still a significant achievement. Can definitely do away the points for the DD.

      I am bit pressed for time at the moment to really give an in depth analysis as to what my thoughts are but I am happy to go with the flow as to what changes are made.
      • Like Like x 1
    15. anthak
      anthak
      actually, it can be done. My memory wasnt quite right. there were some variations I tried that resulted in 100% shooting being way too valuable, but a simple equation of -1 for FGA and 2 fpts for every point, comes out fairly well, its exactly the same as the proposal in the OP actually. except we would need to remove the extra point for FTM because there would already be that extra fpt for every point anyway.
      However, I guess given it is the same outcome, i would prefer to leave the proposal as it is. I think keeping points worth 1 each instead of 2 is better, just for aesthetics and impressions really.
    16. anthak
      anthak
      Ive had a play around with the FTmiss checking differences between -2 and -3.

      Lets look at the two scenarios, -2 or -3...
      At 100% shooting, both will score 2 fpts per bucket.
      The % where its 1 fpt per bucket is 66.6% and 75% respectfully.
      The breakeven % where its zero points scored is 50% and 60% respectfully.
      At 87.5% shooting, they will score 1.71 and 1.57 fpts per point respectfully

      There were about 500 players who shot at least one FT last season. Of those, 392 shot better than 66%, and 279 shot better than 75%.
      460 players shot better than 50%, which is almost all of them, and 428 shot better than 60%.
      Out of all players who shot worse than 55%, only one Mason Plumlee had more than 100 attempts on the season.

      So, if we leave it at -3 for a FTmiss, almost every player (~85% of them) will score positives for their trips to the line overall. I personally think it can be problematic to consider individual trips to the line, especially and-ones etc because they could go awry in the moment, but over time it'll average out, most players will have good games and bad games.

      All of this makes me think its not that different, so just stick with -3... but where it gets interesting is when a player shoots really poorly.
      For instance, at 20% shooting, its a big difference between the two options, with -2FTmiss scoring -6 fpts per point, and -3FTmiss scoring -10 fpts per point... Upon reflection, thats probably too harsh.

      Given its not too different between the two options above 50% shooting, I agree it is probably better to stick with -2 for a FTmiss to avoid the really low scoring from exceptionally poor FT shooting.

      Thoughts?
    17. stowie
      stowie
      I get where @Rad_E_Cool and @anthak are going with the scoring, and happy to run with whatever the majority vote for. Looking at some more players last night, there is a very minimal difference in the final FP scores per game... is it worth all the extra variables to get to the same answer?

      No one number is going to be a perfect representation of the impact a player has had on the game.
    18. Jet
      Jet
      Great work as always Anthak
      I like Steve and Rad e would like small simple changes. Less stats the better IMO, we don’t need to make it about us or ‘AFL’ like scoring cos it’s basketball :) but yes it would be great to reward a players positive impact like SuperCoach.
      The small changes that I would love are:

      -No point deduction for any fouls!
      -No bonus points for dub dubs, trip dubs
      -I personally think blocks should be higher than offensive rebounds.

      I get what you’re saying with offensive boards being really valuable and giving your team another possession, but the reason there isn’t many offensive boards is because by going for them you open yourself up for fast break points. There is a bunch of research, formulas and articles on this if you google it but basically -

      “Looking at the league-average level, the takeaway is this: an NBA team generally improves on offense by about 0.62 points per 100 possessions for each percentage point increase in its offensive rebound rate. This means that if NBA teams were to improve their offensive rebounding from 23% (where it is now) to 30% (where it was a few years ago), they would generally score about 4.3 points more per 100 possessions.”

      So In conclusion I don’t think offensive boards should change from how we currently have them :) the last thing you want is to be cheering on a player in the paint to miss his shot so he gets the put back/tip in.
      • Like Like x 1
    19. smokinjoemisiti
      smokinjoemisiti
      Confirming I will be in for this season, thanks

Share This Page